Optical Character Recognition

From Openmoko

Jump to: navigation, search

Overview

Optical Character Recognition (OCR) for using extracted text from an image or temporary photograph to use in a message, SMS, etc.

GOCR is one of the best open source OCR software there is around. Unfortunately, it drastically lags commercial products in terms of accuracy.

Google's release of HP code named Tesseract is quite good in terms of accuracy. Quick compile, but no gui. I use it for work where having something reasonably accurate that lends itself to scripting is useful. No layout either, but a nice start. GOCR is nowhere near as good.

Retrieved from "http://wiki.openmoko.org/wiki/Optical_Character_Recognition"
Personal tools